Randomized comparison of vaginal self-sampling by standard vs. dry swabs for human papillomavirus testing.

Background: To evaluate if human papillomavirus (HPV) self-sampling (Self-HPV) using a dry vaginal swab is a valid alternative for HPV testing.
Methods: Women attending colposcopy clinic were recruited to collect two consecutive Self-HPV samples: a Self-HPV using a dry swab (S-DRY) and a Self-HPV using a standard wet transport medium (S-WET). These samples were analyzed for HPV using real time PCR (Roche Cobas). Participants were randomized to determine the order of the tests. Questionnaires assessing preferences and acceptability for both tests were conducted. Subsequently, women were invited for colposcopic examination; a physician collected a cervical sample (physician-sampling) with a broom-type device and placed it into a liquid-based cytology medium. Specimens were then processed for the production of cytology slides and a Hybrid Capture HPV DNA test (Qiagen) was performed from the residual liquid. Biopsies were performed if indicated. Unweighted kappa statistics (к) and McNemar tests were used to measure the agreement among the sampling methods.
Results: A total of 120 women were randomized. Overall HPV prevalence was 68.7% (95% Confidence Interval (CI) 59.3-77.2) by S-WET, 54.4% (95% CI 44.8-63.9) by S-DRY and 53.8% (95% CI 43.8-63.7) by HC. Among paired samples (S-WET and S-DRY), the overall agreement was good (85.7%; 95% CI 77.8-91.6) and the κ was substantial (0.70; 95% CI 0.57-0.70). The proportion of positive type-specific HPV agreement was also good (77.3%; 95% CI 68.2-84.9). No differences in sensitivity for cervical intraepithelial neoplasia grade one (CIN1) or worse between the two Self-HPV tests were observed. Women reported the two Self-HPV tests as highly acceptable.
Conclusion: Self-HPV using dry swab transfer does not appear to compromise specimen integrity. Further study in a large screening population is needed.
Trial Registration: ClinicalTrials.gov: NCT01316120.